The resolution performance for a variety of eigenstructure approaches is analyzedand compared.The expressions of SNR thresholds to resolve two uncorrelated equal powersignals for several typical eigenstructure approac...The resolution performance for a variety of eigenstructure approaches is analyzedand compared.The expressions of SNR thresholds to resolve two uncorrelated equal powersignals for several typical eigenstructure approaches are given.The simulation results agree withthe theoretical analysis.展开更多
输入信号的方向向量出现偏差时,最小均方误差算法会出现收敛速度慢、输出性能下降、不稳定等问题.本文针对这些问题,对传统LMS(least mean squares)算法进行了改进,提出了基于Bayesian方法的鲁棒约束LMS算法.该算法利用信号的先验信息...输入信号的方向向量出现偏差时,最小均方误差算法会出现收敛速度慢、输出性能下降、不稳定等问题.本文针对这些问题,对传统LMS(least mean squares)算法进行了改进,提出了基于Bayesian方法的鲁棒约束LMS算法.该算法利用信号的先验信息对实际信号方向向量进行估计,有效地抑制了方向向量偏差的影响,并提高了系统的鲁棒性.阵列输出的信干噪比得到了改善,更加接近最优值.仿真实验验证了该算法的有效性和可行性.展开更多
文摘The resolution performance for a variety of eigenstructure approaches is analyzedand compared.The expressions of SNR thresholds to resolve two uncorrelated equal powersignals for several typical eigenstructure approaches are given.The simulation results agree withthe theoretical analysis.
文摘输入信号的方向向量出现偏差时,最小均方误差算法会出现收敛速度慢、输出性能下降、不稳定等问题.本文针对这些问题,对传统LMS(least mean squares)算法进行了改进,提出了基于Bayesian方法的鲁棒约束LMS算法.该算法利用信号的先验信息对实际信号方向向量进行估计,有效地抑制了方向向量偏差的影响,并提高了系统的鲁棒性.阵列输出的信干噪比得到了改善,更加接近最优值.仿真实验验证了该算法的有效性和可行性.